Intelligent Classification Method Research Articles

Multi-Feature Extraction and Selection Method to Diagnose Burn Depth from Burn Images

Burn wound depth is a significant determinant of patient treatment. Typically, the evaluation of burn depth relies heavily on the clinical experience of doctors. Even experienced surgeons may not achieve high accuracy and speed in diagnosing burn depth. Thus, intelligent burn depth classification is useful and valuable. Here, an intelligent classification method for burn depth based on machine learning techniques is proposed. In particular, this method involves extracting color, texture, and depth features from images, and sequentially cascading these features. Then, an iterative selection method based on random forest feature importance measure is applied. The selected features are input into the random forest classifier to evaluate this proposed method using the standard burn dataset. This method classifies burn images, achieving an accuracy of 91.76% when classified into two categories and 80.74% when classified into three categories. The comprehensive experimental results indicate that this proposed method is capable of learning effective features from limited data samples and identifying burn depth effectively.

Intelligent classification and identification method for Conger myriaster freshness based on DWG‐YOLOv8 network model

Intelligent classification and identification method for <i>Conger myriaster</i> freshness based on DWG‐YOLOv8 network model

A sensitivity indicator screening and intelligent classification method for the diagnosis of T2D-CHD.

The prevalence of Type 2 Diabetes Mellitus (T2D) and its significant role in increasing Coronary Heart Disease (CHD) risk highlights the urgent need for effective CHD screening within this population. Despite current advancements in T2D management, the complexity of cardiovascular complications persists. Our study aims to develop a comprehensive CHD screening model for T2D patients, employing multimodal data to improve early detection and management, addressing a critical gap in clinical practice. We analyzed data from 699 patients, including 471 with CHD (221 of these also had T2D) and a control group of 228 without CHD. Employing strict diagnostic criteria, we conducted significance testing and multivariate analysis to identify key indicators for T2D-CHD diagnosis. This led to the creation of a neural network model using 21 indicators and a logistic regression model based on an 8-indicator subset. External validation was performed with an independent dataset from an additional 212 patients to confirm the models' generalizability. The neural network model achieved an accuracy of 90.7%, recall of 90.78%, precision of 90.83%, and an F-1 score of 0.908. The logistic regression model demonstrated an accuracy of 90.13%, recall of 90.1%, precision of 90.22%, and an F-1 score of 0.9016. External validation reinforced the models' reliability and effectiveness in broader clinical settings. Our AI-driven diagnostic models significantly enhance early CHD detection and management in T2D patients, offering a novel, efficient approach to addressing the complex interplay between these conditions. By leveraging advanced analytics and comprehensive patient data, we present a scalable solution for improving clinical outcomes in this high-risk population, potentially setting a new standard in personalized care and preventative medicine.

Intelligent Recognition of Valid Microseismic Events Based on Self-supervised Learning

The identification of valid microseismic events is a crucial stage in the processing of microseismic data and plays a central role in subsequent monitoring and early warning of rockburst disasters. However, obtaining microseismic data is difficult, and manually identifying effective microseismic signals is even more challenging. Therefore, exploring high-precision intelligent classification methods with limited labeled data is necessary. In this study, we applied a self-supervised learning method based on contrastive learning to classify small samples of labeled data. Firstly, use large unlabeled data to pre-train SimCLR (a simple contrastive learning framework) for transfer learning to classify small sample labeled data. The classification accuracy was 87.14%, an increase of 11.3% compared to the supervised-deep-learning method. Furthermore, we directly applied the pre-trained SimCLR to the microseismic data of various projects, achieving a classification accuracy of 86.70%. This demonstrates strong classification performance and generalization and provides valuable technical support for the subsequent monitoring and warning of rockburst disasters.

Novel Intelligent Methods for Channel Path Classification and Model Determination Based on Blind Source Signals

In this paper, the urban signal propagation characteristics based on the location of blind sources are investigated. To address the issue of blind electromagnetic radiation sources in complex urban environments, intelligent methods for propagation channel path classification, and model determination are brought forth based on field test data. The intelligent classification method distinguishes between the Line-of-Sight (LoS) path channel and a direct path, the LoS multipath channel with a direct path and other multiple paths, and the Non-Line-of-Sight (NLoS) multipath channel without a direct path from the source to the test point. The modeling aspect determines the model type to which the received signal belongs based on the statistical model derived from the tested data of a specific source. A validation measurement system was constructed for the FM broadcasting band, and validation campaigns were conducted in the city of Wuhan. The process and analysis of the data using this method demonstrate the accurate distinction of the different propagation path channels and models and involve the construction of a statistical model for the FM band in Wuhan’s urban area.

An intelligent identification and classification system of decoration waste based on deep learning model

Efficient sorting and recycling of decoration waste are crucial for the industry's transformation, upgrading, and high-quality development. However, decoration waste can contain toxic materials and has greatly varying compositions. The traditional method of manual sorting for decoration waste is inefficient and poses health risks to sorting workers. It is therefore imperative to develop an accurate and efficient intelligent classification method to address these issues. To meet the demand for intelligent identification and classification of decoration waste, this paper applied the deep learning method You Only Look Once X (YOLOX) to the task and proposed an identification and classification framework of decoration waste (YOLOX-DW framework). The proposed framework was validated and compared using a multi-label image dataset of decoration waste, and a robot automatic sorting system was constructed for practical sorting experiments. The research results show that the proposed framework achieved a mean average precision (mAP) of 99.16 % for different components of decoration waste, with a detection speed of 39.23 FPS. Its classification efficiency on the robot sorting experimental platform reached 95.06 %, indicating a high potential for application and promotion. This provides a strategy for the intelligent detection, identification, and classification of decoration waste.

Intelligent monitoring of milling tool wear based on milling force coefficients by prediction of instantaneous milling forces

The spindle current of a machine tool contains a wealth of machining information and by analyzing its signal characteristics wear monitoring of milling tools can be achieved. Most current research in tool wear monitoring has been achieved by intelligent classification methods, but such methods are unable to identify wear states when cutting parameters change. To address this challenge, the paper proposes a new tool wear monitoring framework that identifies the wear state of milling cutters under different cutting parameters. The first step is to establish the mapping relationship between the spindle current signal and the milling force signal through neural networks under a variety of cutting parameters. Based on the mapping property between two signals, a TCN-LSTM based neural network prediction model is proposed to achieve accurate prediction of milling forces by spindle current signal. Various cutting information is then calculated based on the predicted milling forces and PLC communication, and a method is proposed to identify the machining status of the tool: the milling force coefficient monitoring method. The method reflect the wear status of the milling cutter by monitoring the range of threshold values, making the identification of the tool wear state without interference from variable spindle speeds, feeds and depths of cut. The proposed method provides a new solution for tool wear monitoring, breaking the inherent thinking of intelligent tool wear classification, and the experimental results prove the effectiveness and generality of the proposed method.

Model-Aided Approach for Intelligent Fault Detection System for SF6 High-Voltage Circuit Breaker With Spring Operating Mechanism

Supervised machine learning (ML) methods have been increasingly applied in the fault detection of components. However, they relied on the data set and data acquisition under faulty conditions, which is time-consuming with the risk of component damage. This paper presents a model to rectify this point through simulation of the spring drive operating mechanism as the main origin of major failures for a real case, i.e., a 72.5 kV, SF6, EDF high voltage circuit breaker (HVCB). The model is simulated usingmechanical analysis software such asADAMSand SOLIDWORKS to capture the travel curve (TC) as the primary signal for diagnosing HVCB mechanical faults. The model has been verified against the measured TC. Subsequently, a comprehensive database is generated consisting of various mechanical defects in addition to the healthy operation of HVCB. Afterward, different ML algorithms have been trained and compared. The results are indicated that support vector machines with a Gaussian kernel and decision tree are the most effective models for detecting operating mechanisms in trip/close operations. In addition to solving the data collection problem, intelligent classification methods were employed in this research to increase fault detection accuracy.

The hesitant fuzzy BiRNN based on twice-cycle mechanism and its intelligent applications

Existing intelligent classification methods could be inefficient to deal with the hybrid environments including hesitant fuzzy information and real numbers. With respect to this real issue, in this study, we propose some new intelligent methods to achieve deep learning and intelligent classification under this hybrid environment. To do this, we construct the hesitant fusion bidirectional recurrent neural network (HF-BiRNN) based on the hesitant fusion mechanism. Then, the twice-cycle mechanism is designed, which includes the extension mechanism and the decomposition-reorganization mechanism, to fully utilize the original data and optimize the classification results. Meanwhile, the overlap degree algorithm is constructed to filter the optimal outputs. After that, we further propose the hesitant expansion BiRNN (HE-BiRNN) by combining with the twice-cycle mechanism, overlap degree algorithm, and BiRNN. Lastly, these new methods are used to the problems of driving route classification and red wine quality assessment. The derived optimal results and comparison analysis fully show the effectiveness and feasibility of the new proposed mechanisms and developed models.

Electrocardiogram classification algorithm based on CvT-13 and multimodal image fusion

Electrocardiogram (ECG) signal is an important basis for the diagnosis of arrhythmia and myocardial infarction. In order to further improve the classification effect of arrhythmia and myocardial infarction, an ECG classification algorithm based on Convolutional vision Transformer (CvT) and multimodal image fusion was proposed. Through Gramian summation angular field (GASF), Gramian difference angular field (GADF) and recurrence plot (RP), the one-dimensional ECG signal was converted into three different modes of two-dimensional images, and fused into a multimodal fusion image containing more features. The CvT-13 model could take into account local and global information when processing the fused image, thus effectively improving the classification performance. On the MIT-BIH arrhythmia dataset and the PTB myocardial infarction dataset, the algorithm achieved a combined accuracy of 99.9% for the classification of five arrhythmias and 99.8% for the classification of myocardial infarction. The experiments show that the high-precision computer-assisted intelligent classification method is superior and can effectively improve the diagnostic efficiency of arrhythmia as well as myocardial infarction and other cardiac diseases.

Non-negative wavelet matrix factorization-based bearing fault intelligent classification method

There are more and more bearing fault types under considering the fault location and degree, and the corresponding fault classification task is becoming increasingly heavy. Raw signals that have not been processed or simply processed are directly input into convolutional neural network (CNN) for classification, resulting in poor classification performance. Aiming at this issue, a time–frequency joint metric feature extraction technique named non-negative wavelet matrix factorization (NWMF) is developed to extract more effective features by comprehensively leveraging the advantages of continuous wavelet transform and non-negative matrix factorization. Based on the NWMF and CNN, an effective intelligent diagnosis framework is constructed to detect bearing fault. In the proposed framework, based on the NWMF, a non-negative basic matrix with smaller size is calculated from the original time–frequency spectrum and it includes bearing fault-related internal core information. In addition, a novel CNN is developed to identify locations and sizes of fault bearing based on the calculated internal core information. For verifying the effectiveness of the proposed framework in handling heavier tasks, the types of bearing faults in the experiments are set up to 15, the results and comparative analysis reveal that the feasibility and superiority of the proposed method are much better than the other traditional machine learning methods and original deep learning methods, such as the support vector machine, random forest and residual neural network.

A novel fault classification feature extraction method for rolling bearing based on multi-sensor fusion technology and EB-1D-TP encoding algorithm

To improve the accuracy of bearing fault diagnosis in a multisensor monitoring environment, it is necessary to obtain more accurate and effective fault classification features for bearings. Accordingly, a bearing fault classification feature extraction method based on multisensor fusion technology and an enhanced binary one-dimensional ternary pattern (EB-1D-TP) algorithm were proposed in this study. First, an optimal equalization weighting algorithm was established to realize high-precision fusion of bearing signals by introducing an optimal equalization factor and determining the theoretical optimal equalization factor value. Second, an enhanced binary encoding method similar to balanced ternary encoding was developed, which increases the difference between the different fault features of the bearing. Finally, the new sequence obtained after encoding was used as the input to a support vector machine to classify and diagnose the faults of the rolling bearing. The experimental results show that the algorithm can significantly improve the accuracy and speed of rolling-bearing fault classification. Combining fusion-encoding features with other intelligent classification methods, the classification results were improved.

A novel discrete learning-based intelligent methodology for breast cancer classification purposes

Classification is one of the most significant subfields of data mining that has been successfully applied to various applications. The literature has expended substantial effort to present more efficient and accurate classification models. Despite the diversity of the proposed models, they were all created using the same methodology, and their learning processes ignored a fundamental issue. In all existing classification model learning processes, a continuous distance-based cost function is optimized to estimate the unknown parameters. The classification problem's objective function is discrete. Consequently, applying a continuous cost function to a classification problem with a discrete objective function is illogical or inefficient. This paper proposes a novel classification methodology utilizing a discrete cost function in the learning process. To this end, one of the most popular intelligent classification models, the multilayer perceptron (MLP), is used to implement the proposed methodology. Theoretically, the classification performance of the proposed discrete learning-based MLP (DIMLP) model is not dissimilar to that of its continuous learning-based counterpart. Nevertheless, in this study, to demonstrate the efficacy of the DIMLP model, it was applied to several breast cancer classification datasets, and its classification rate was compared to that of the conventional continuous learning-based MLP model. The empirical results indicate that the proposed DIMLP model outperforms the MLP model across all datasets. The results demonstrate that the presented DIMLP classification model achieves an average classification rate of 94.70 %, a 6.95 % improvement over the classification rate of the traditional MLP model, which was 88.54 %. Therefore, the classification approach proposed in this study can be utilized as an alternative learning process in intelligent classification methods for medical decision-making and other classification applications, particularly when more accurate results are required.

Anti-local occlusion intelligent classification method based on MobileNet for hazardous waste

Anti-local occlusion intelligent classification method based on MobileNet for hazardous waste

Anti-local occlusion intelligent classification method based on MobileNet for hazardous waste

Anti-local occlusion intelligent classification methods based on MobileNet and VTM for hazardous waste are investigated in this paper. Three image datasets with ten kinds of hazardous waste and 5,000.0 samples are constructed, which include the image dataset with without occlusion, the image dataset with 15% occlusion, and the image dataset with random occlusion. Based on them, the MobileNet and VTM intelligent classification model are constructed, trained, and tested, respectively. It can be seen from testing results that the classification accuracies of VTM and MobileNet are very high for the image dataset with without occlusion. But as occlusion area on images go up or randomly changes, the classification accuracies of VTM and MobileNet go down for 15% and random occlusion cases. The testing results show that classification accuracy of MobileNet model is better than that of VTM model for hazardous waste with or without occlusion.

A Novel Brain Tumor Classification Algorithm using SVM Classifier

This study provides an intelligent classification method for distinguishing between abnormal and normal MRI brain images. Medical pictures like MRI, ECG, and CTscan pictures are vital tools for accurately diagnosing human disease. Whenever a lot of MRIs need to be examined, traditional approach of manual tumour analysis, which relies on visual examination by a physician and radiologist, might lead to inaccurate classification. To remove human mistakes, a proposal is made for an intelligent classification system that responds to the essentials of image classification. Brain tumours are one of the primary causes of human mortality. If a tumor is diagnosed appropriately at an early stage, the chances of survival can be improved. The human brain is studied using the MRI method. The acronym MRI stands for magnetic resonance imaging. In this study, classification strategies based on Support Vector Machines (SVM) are proposed and used to brain imaging categorization. In this research, grayscale, symmetry, and texture features are utilised to extract features from MRI images. The fundamental objective of this study is to offer a decent classification result (improved accuracy and lower error rate) to detect MRI brain tumours with help of SVM. Keywords— Brain tumor, Classification, SVM, MRI.

A novel grid-based many-objective swarm intelligence approach for sentiment analysis in social media

Sentiment analysis is a field of study that analyses people's opinions, evaluations, feelings, ratings, sentiments, and attitudes towards entities such as products, organizations, individuals, services, topics, titles, events, and qualifications. Studies on sentiment analysis problems in social media have generally adopted intelligent classification methods. However, there are conflicting and contradictory objectives to simultaneously optimize, and active research continues into developing a more effective analysis model in terms of many metrics in order to achieve effective usage. This study considers sentiment analysis as a many-objective optimization problem for the first time. For this purpose, it first proposes a Grid-based Adaptive Many-Objective Grey Wolf Optimizer (GAM-GWO) based on the Grey Wolf Optimizer algorithm. Then, it adapts this proposed method for the sentiment analysis problem in order to obtain more successful results in terms of different metrics. The study tests the performance of the proposed approach with three different data sets. Experimental results show that GAM-GWO can achieve non-dominated and competitive results in all data set classes.

Research on an Intelligent Identification and Classification Method of Complex Holes in Triangle Meshes for 3D Printing.

In triangular mesh models, the repair of complex hole poses a difficult problem, which always causes serious repair defects. Therefore, it is needed to develop an intelligent identification and classification method of complex holes to reduce repair difficulties. First, the topological structure of the complex hole is studied and all the holes are divided into single holes and continuous holes depending on whether there are intersection points. Second, to tackle the nesting and connecting of complex continuous holes, a decomposition method of multiply connected domains based on intersection points is proposed to partition or reconstruct complex continuous holes into single holes. Based on the different geometric structures, single holes are classified into five common hole types and a corresponding identification method of single holes is presented. Finally, an experiment is carried out to verify the repair quality and efficiency of the proposed method. Compared with Geomagic software, the proposed method can automatically identify and partition complex holes with fewer defects and similar efficiency. It can reduce the difficulties of repairing complex holes and enable the repair of complex holes based on existing methods. It is shown that the method can be applied to complex hole repair of 3D printing models without the participation of technicians.

Intelligent Classification Method of Archive Data Based on Multigranular Semantics.

With the rapid development of information technology, the amount of data in various digital archives has exploded. How to reasonably mine and analyze archive data and improve the effect of intelligent management of newly included archives has become an urgent problem to be solved. The existing archival data classification method is manual classification oriented to management needs. This manual classification method is inefficient and ignores the inherent content information of the archives. In addition, for the discovery and utilization of archive information, it is necessary to further explore and analyze the correlation between the contents of the archive data. Facing the needs of intelligent archive management, from the perspective of the text content of archive data, further analysis of manually classified archives is carried out. Therefore, this paper proposes an intelligent classification method for archive data based on multigranular semantics. First, it constructs a semantic-label multigranular attention model; that is, the output of the stacked expanded convolutional coding module and the label graph attention module are jointly connected to the multigranular attention Mechanism network, the weighted label output by the multigranularity attention mechanism network is used as the input of the fully connected layer, and the output value of the fully connected layer used to map the predicted label is input into a Sigmoid layer to obtain the predicted probability of each label; then, the model for training: use the multilabel data set to train the constructed semantic-label multigranularity attention model, adjust the parameters until the semantic-label multigranularity attention model converges, and obtain the trained semantic-label multigranularity attention model. Taking the multilabel data set to be classified as input, the semantic-label multigranularity attention model after training outputs the classification result.

Automatic Classification and Identification of Road Garbage Images and Evaluation of Environmental Health Based on UNet++

As Covid-19 plagues the world, a clean environment helps to control the factors and risks that threaten health, and curb the spread of the epidemic. However, the quality evaluation of environmental health faces some problems and challenges in actual management and practice. Firstly, the classification, identification, and quantification of road garbage are mainly done manually, because of the diversity of road garbage, as well as their sharp differences in geometry, color, and texture. Secondly, it is labor-intensive to manually manage the large operation areas on the wide urban roads. Thirdly, the accuracy of statistical indices is affected by the time-varying road environment, making the quality evaluation of environmental health untimely and inaccurate. To solve these problems, this paper proposes an intelligent image classification and evaluation method for urban environmental health. Specifically, an environmental garbage recognition and semantic segmentation approach was designed based on UNet++, and combined with the vehicle-mounted machine vision system to automatically identify the typical targets among the road waste control indices. Next, an image attention quantitative evaluation method was developed based on the eye tracking analyzer, and the quantified attention was fused with the statistical features for road garbage classification, forming an attention-based evaluation method for environmental quality. The proposed approach supports the automatic recognition and semantic segmentation of the garbage on urban roads, and realizes the identification of complex targets in different scenes through transfer learning. In addition, the attention-based evaluation method for environmental quality provides environmental management departments with visual basis for quantitative decision-making.